CN115114714B - Municipal engineering excavation and filling auxiliary method based on excavation form measurement - Google Patents

Abstract

The invention relates to the technical field of excavation and filling engineering, in particular to a municipal engineering excavation and filling auxiliary method based on excavation form measurement, which is used for acquiring a compaction state stability index, the typical degree of an excavation volume increment value and a Fourier descriptor of an excavation section form in excavation work so as to calculate the work special case degree of the excavation work, and acquiring the section form characteristic of the excavation work by the Fourier descriptor; forming a state feature vector of excavation work by the compaction state stability index, the typical degree of the excavation volume increment value, the special case degree of work and the section morphological feature; and forming a normal sample set and an abnormal sample set by the state feature vectors of a plurality of continuous excavation works, training an Adaboost two-classifier, and detecting the abnormal state of the excavation works by using the trained Adaboost two-classifier. The invention can quickly and accurately carry out abnormity early warning so as to conveniently find problems in time and play an auxiliary role in excavation.

Description

Municipal engineering excavation and filling auxiliary method based on excavation form measurement

Technical Field

The invention relates to the technical field of excavation and filling engineering, in particular to a municipal engineering excavation and filling auxiliary method based on excavation form measurement.

Background

The excavation and filling refers to excavation and filling, and the excavation refers to the volume of earth and stones excavated from the original ground to the surface of the roadbed when the surface of the roadbed is lower than the original ground; the filling is the volume of earth and stones filled from the original ground to the surface part of the roadbed when the surface of the roadbed is higher than the original ground. The excavation and filling is an important early leveling process in the building engineering. When the excavation process is abnormal, the general industry people carry out problem preliminary judgment on the current excavation condition against the information of the section form, the judgment mode needs to be completed with abundant experience, and depends on subjective consciousness, so that the problem is difficult to continuously observe, and meanwhile, certain problems can not be accurately distinguished due to too many section form characteristics.

Disclosure of Invention

In order to solve the technical problems, the invention aims to provide a municipal engineering excavation and filling auxiliary method based on excavation form measurement, and the adopted technical scheme is as follows:

taking a working process corresponding to a set excavation and filling path as excavation work, acquiring an excavation and filling compaction square coefficient and an excavation volume increment value in the current excavation work based on a set sampling frequency, and respectively and correspondingly obtaining an excavation and filling compaction square coefficient sequence and an excavation volume increment value sequence; calculating a compaction state stability index from the sequence of excavation and packing compactor coefficients, and calculating a typical degree of excavation volume increment values from the sequence of excavation volume increment values; acquiring the excavation section form of the current excavation work, and acquiring a Fourier descriptor of the excavation section form according to the point cloud data of the excavation section form;

calculating the working special case degree of the current excavation work by combining the Fourier descriptor, the excavation and filling compaction square coefficient sequence and the compaction state stability index; taking the low-frequency component in the Fourier descriptor as the section morphological characteristic of the current excavation work; forming the compaction state stability index, the typical degree of the excavation volume increment value, the working special case degree and the section morphological characteristics into a state characteristic vector of the current excavation work;

acquiring the state characteristic vectors of a plurality of continuous excavation works, and clustering all the state characteristic vectors according to the difference value between the work special case degrees to obtain a normal sample set and an abnormal sample set; and training an Adaboost two-stage classifier by using the normal sample set and the abnormal sample set to obtain a strong classifier of the Adaboost two-stage classifier, and detecting the abnormal state of excavation work by using the strong classifier.

Further, the calculation formula of the compaction state stability index is as follows:

wherein the content of the first and second substances,

is a compaction state stability index;

is an averaging function;

is a function of absolute values;

for the first in the sequence of fill and compaction square coefficients

The individual excavation, filling and compaction square coefficient;

for the first in the sequence of fill and compaction square coefficients

The individual excavation, filling and compaction square coefficient;

is a constant.

Further, the method for obtaining the typical degree of the excavation volume increment value comprises the following steps:

respectively calculating the square of the difference between each excavation volume increment value in the excavation volume increment value sequence and the reference excavation volume to obtain the mean value of the square of the difference, and taking the product of the mean value and a correction coefficient as a difference index; the reference excavation volume refers to the excavation volume before compaction;

and optimizing the difference index by using a hyperbolic tangent function to obtain the typical degree of the excavation volume increment value.

Further, the calculation formula of the degree of the specific working case is as follows:

wherein the content of the first and second substances,

is the degree of the special case of the work; cosine (—) is a Cosine similarity function; mean (—) is a mean function; max is a function for solving the maximum value;

is a reference fourier descriptor;

is the Fourier descriptor;

is the compaction state stability indicator;

and the sequence of the excavation and filling compaction square coefficients is obtained.

Further, the method for acquiring the normal sample set and the abnormal sample set includes:

respectively calculating the difference value between the degrees corresponding to the special cases of the work in any two excavation works, taking the difference value as a sample distance, and performing DBSCAN clustering on all state feature vectors based on the sample distance to obtain a plurality of cluster clusters and isolated points;

and forming the state characteristic vectors corresponding to the isolated points into the abnormal sample set, and forming the state characteristic vectors in all the clustering clusters into the normal sample set.

Further, the abnormal sample set is optimized, and the optimization method comprises the following steps:

counting the working special case degrees in the abnormal sample set to obtain a median of the working special case degrees, obtaining target state feature vectors which are larger than the median in the abnormal sample set, randomly combining elements in all the target state feature vectors to obtain a plurality of new state feature vectors, and placing the new state feature vectors into the abnormal sample set.

Further, the method for detecting the abnormal state of the excavation work by using the strong classifier comprises the following steps:

respectively obtaining historical response values of a plurality of historical excavation works according to the strong classifiers, and obtaining response value thresholds according to the historical response values;

and respectively acquiring a real-time response value of real-time excavation work and a response value of continuous K times of excavation work before the real-time excavation work according to the strong classifier, and performing excavation abnormity early warning when the real-time response value and the response value are lower than the response value threshold value.

The embodiment of the invention at least has the following beneficial effects: (1) The method comprises the steps of collecting excavation filling compaction square coefficients, excavation volume increment values and excavation section forms in excavation work to obtain state feature vectors of the excavation work, training an Adaboost two-classifier by utilizing the state feature vectors of a plurality of continuous excavation works to obtain the Adaboost two-classifier for monitoring the excavation work state, and achieving intelligent monitoring.

(2) Utilize unmanned aerial vehicle to carry out real-time data acquisition to excavation work to in time obtain the state eigenvector of excavation work, and obtain the response value that excavation work corresponds through the good Adaboost two classifiers of training with the state eigenvector of continuous excavation work, can be fast accurate carry out unusual early warning according to the response value, in order to make things convenient for in time finding the problem, play the secondary action to the excavation, reduce anti-worker loss and secondary loss.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a flowchart illustrating steps of a municipal engineering cut and fill assisting method based on cut figure measurement according to an embodiment of the present invention.

Detailed Description

In order to further explain the technical means and effects of the present invention adopted to achieve the predetermined invention purpose, the following detailed description, the specific implementation manner, the structure, the characteristics and the effects thereof, of the municipal engineering excavation and filling auxiliary method based on excavation form measurement according to the present invention, with reference to the accompanying drawings and the preferred embodiments. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The concrete scheme of the municipal engineering excavation and filling auxiliary method based on excavation form measurement provided by the invention is specifically described below with reference to the accompanying drawings.

Referring to fig. 1, a flow chart of steps of a municipal engineering excavation and filling assistance method based on excavation form measurement according to an embodiment of the present invention is shown, the method including the following steps:

step S001, taking a working process corresponding to a set excavation and filling path as excavation work, acquiring an excavation and filling compaction square coefficient and an excavation volume increment value in the current excavation work based on a set sampling frequency, and respectively and correspondingly obtaining an excavation and filling compaction square coefficient sequence and an excavation volume increment value sequence; calculating a compaction state stability index from the excavation and filling compaction square coefficient sequence, and calculating a typical degree of excavation volume increment value from the excavation volume increment value sequence; and acquiring the excavation section form of the current excavation work, and acquiring a Fourier descriptor of the excavation section form according to the point cloud data of the excavation section form.

Specifically, compaction is a very important process of excavation work, and based on the airborne laser radar of the unmanned aerial vehicle, the height difference before and after compaction can be calculated by combining oblique photography results, and the height difference is used as an excavation and filling compaction square coefficient.

Under normal compaction, the cut and fill compactor square coefficient is stabilized at a value, if smallIn a stable state, the compaction operation can be proved to be irregular to a certain extent, and at the moment, the possibility of abnormal form can be generated, so that the quality of the whole excavation and filling is influenced, therefore, the data sampling is carried out on the compaction condition of the excavation work, and the coefficient sequence of the excavation and filling compaction square in the excavation work is obtained

Wherein, in the step (A),

is a first

Fill pack coefficient of subsampling.

It should be noted that the sampling method is as follows: and acquiring mileage once every 5m based on the central axis of the excavation and filling path, wherein the working process corresponding to the 50m excavation and filling path is excavation work.

For the obtained coefficient sequence of the excavation, filling and compaction square

Carrying out pretreatment to determine the stability of the compaction state in excavation work, wherein the method comprises the following steps: respectively calculating the ratio of any two adjacent excavation and filling compaction square coefficients in the excavation and filling compaction square coefficient sequence, and calculating the compaction state stability index in excavation work based on the ratio

：

Wherein, the first and the second end of the pipe are connected with each other,

the average function is used for acquiring the average value of the processed data;

is an absolute value function and is used for taking the processed data into a non-negative region;

for filling out the second in the series of square coefficients

The individual excavation, filling and compaction square coefficient;

for filling out the second in the series of square coefficients

The individual excavation, filling and compaction square coefficient;

is a constant.

It should be noted that, when the coefficients of two adjacent excavation and filling compaction squares are the same, the processed index is 1, and if the two coefficients are not equal, the index is less than 1, and the stable condition of the compaction state can be directly judged through the magnitude of the index function; when the processed average value in the whole record is 1, the sequence of the excavation, filling and compaction square coefficients has no fluctuation, namely, the performance is stable.

When the volume increase value of the excavation is low, the quality of compaction may be poor, and therefore, the voxel volume of the road surface difference before and after excavation is calculated based on the oblique photography of the unmanned aerial vehicle to serve as the volume increase value of the excavation, wherein the calculation of the voxel volume uses a known technology, which is not described herein in detail.

The sampling mode of voxel volume calculation is consistent with the excavation and filling compaction square coefficient, namely, based on the central axis of an excavation and filling path, the mileage of every 5m is collected, the working process corresponding to the 50m excavation and filling path is one excavation work, and therefore, an excavation volume increment value sequence in the excavation work can be obtained

Wherein, in the step (A),

is a first

A sub-sampled squared volume delta value.

Similarly, for the sequence of the obtained excavation volume increment values

Carrying out pretreatment: respectively calculating the difference between each excavation volume increment value in the excavation volume increment value sequence and the reference excavation volume under engineering experience, and calculating a difference index by combining all the differences corresponding to the excavation volume increment value sequence, so that the difference index

The calculation formula of (2) is as follows:

wherein the content of the first and second substances,

for the pre-compaction excavation volume, the implementer estimates a 5m excavation volume as the excavation volume based on excavation requirements

；

For correcting the coefficient, the volume increment value of the excavation in normal time is reduced to

In (i), i.e.

The inverse of the empirically maximum volume increment value is preferred

。

Note that the difference indicator is used when the difference between each of the sequence of dipper volume increment values and the reference dipper volume under engineering experience is greater

The larger the difference, the smaller the difference, the difference index

Smaller and therefore able to pass the disparity index

To reflect the compaction quality in the excavation work.

Furthermore, in order to obtain the valid interval for the voxel volume calculation range, the difference index is adjusted

Optimizing to obtain typical degree of volume increment value

Then, the optimization processing method is as follows:

wherein the content of the first and second substances,

is a hyperbolic tangent function processed in a section

In the interior of said container body,

the value is about 1, and conversely,

close to 0.

When the scheme is used for collecting the excavation section form, the default excavation work is carried out as expected, and the influence of extreme factors on the generated excavation section form is not considered. The collection of excavation section forms is carried out in a mode of model slicing, a concrete recording mode is determined by an implementer according to an engineering tool (such as ArcGIS), a sampling rate is determined according to the precision of an unmanned aerial vehicle airborne laser radar, according to experience, the section height value of excavation is recommended to be not less than 80 sample points, then the unmanned aerial vehicle is used for collecting the excavation section forms of one excavation work, point cloud data of the excavation section forms are obtained, and one excavation work corresponds to one excavation section form.

The form can show that whether the current excavation condition is the same as the rest excavation conditions due to the judgment error of an excavator or human eyes in the excavation construction process, so that the abnormal condition of the engineering construction process can be determined through the section form information, and the excavation section form is preprocessed: taking the point cloud data of the excavation section form as an outline, and further acquiring a Fourier descriptor of the outline

Wherein the Fourier descriptor

Are ordered by frequency magnitude.

Sequence of fill and compaction square coefficients for a excavation operation

And a stable index of compacted state

Sequence of volume increment values

Square digging bodyTypical degree of product increment value

And Fourier descriptor of the cut section form

Characteristic data of the excavation work is composed.

Step S002, calculating the working special case degree of the current excavation work by combining the Fourier descriptor, the excavation and filling compaction square coefficient sequence and the compaction state stability index; taking a low-frequency component in the Fourier descriptor as a section morphological feature of the current excavation work; and forming the compaction state stability index, the typical degree of the excavation volume increment value, the special case degree of the work and the section morphological characteristics into a state characteristic vector of the current excavation work.

Specifically, when the excavation work is abnormal, the excavation section form and the excavation filling and compaction square coefficient can obviously change, so that the degree of the special case of the excavation work is determined according to the characteristic data of the excavation work

：

Wherein, cosine (, x) is a Cosine similarity function, and is used for comparing the similarity degree between two vectors; mean (—) is a mean function; max is a function for solving the maximum value;

and the reference Fourier descriptor is used as a reference Fourier descriptor when the excavation work is in normal work.

It should be noted that when the fourier descriptor of the real-time excavation work is closer to the reference fourier descriptor of the normal work, the cosine similarity is larger, and the degree of the corresponding work special case is smaller; the greater the stability of the excavation filling compaction square coefficient of excavation work is, namely the greater the compaction state stability index is, the smaller the corresponding work special case degree is; if the mean value of the excavation, filling and compaction square coefficients is obviously different from the maximum value, the ratio is smaller, and the corresponding working special case degree is larger.

Because the fourier descriptors in the feature data of the excavation work contain more information and part of the information is errors or small bumps and pits, the high-frequency components of the fourier descriptors are useless, and the fourier descriptors are processed: and extracting low-frequency components in the Fourier descriptor, forming the low-frequency components into a section morphological feature S of excavation work, and expressing the low-frequency components by fewer vectors by using the low-frequency components so as to reduce subsequent calculation amount.

Preferably, because the number of stages of the fourier descriptors is not fixed, the first 5 low-frequency components are selected to form the cross-sectional morphological feature S in the embodiment of the present invention.

The purpose of representing the morphological characteristics of the cross section by using the low-dimensional vector is to reduce the calculation amount when the Adaboost two classifier is established, improve the judgment performance and avoid errors caused by foreign matters on the excavation bottom surface.

Stabilization of compaction state of an excavation

Typical degree of volume increase value of excavation

Degree of special case of work

And the section morphological characteristics S form a state characteristic vector X = { Q, P, U, S } corresponding to excavation work.

Step S003, obtaining state characteristic vectors of a plurality of continuous excavation works, and clustering all the state characteristic vectors according to the difference value between the work special case degrees to obtain a normal sample set and an abnormal sample set; and training the Adaboost two classifiers by using the normal sample set and the abnormal sample set to obtain a strong classifier of the Adaboost two classifiers, and detecting the abnormal state of excavation work by using the strong classifier.

Specifically, since the abnormal condition of a single excavation work is relatively single and the probability of the abnormal condition is relatively low, the state feature vectors of a plurality of continuous excavation works are obtained by the methods of step S001 and step S002.

And respectively calculating the difference value between the special case degrees of corresponding work in any two excavation works, taking the difference value as a sample distance, and carrying out DBSCAN clustering on all state feature vectors based on the sample distance, wherein the search radius eps in the DBSCAN clustering is defaulted to be 0.5, and the minimum value minpts in the clustering is set to be 4, so that a plurality of clustering clusters and isolated points are obtained.

In the embodiment of the invention, the state characteristic vectors corresponding to the isolated points are used as abnormal samples to further obtain an abnormal sample set, and the state characteristic vectors in all the clustering clusters are used as normal samples to further obtain a normal sample set.

And taking the state feature vector as input data of an Adaboost two-classifier, and training the Adaboost two-classifier by using the abnormal sample set and the normal sample set to determine a strong classifier of the Adaboost two-classifier.

Because the number of the abnormal samples is small, the states of all weak classifiers in the Adaboost two classifiers cannot be effectively constrained, and the abnormal samples are expanded according to the state feature vectors in the abnormal sample set: since the abnormal condition can be determined more obviously under the condition of larger work special case degree U, the work special case degree U in the abnormal sample set is counted to obtain a median value of the work special case degree

For a value greater than in the abnormal sample set

The state feature vectors are expanded, namely the compaction state stability indexes in the state feature vectors

Volume increase of excavationTypical degree of value

Degree of special case of work

And randomly combining the abnormal state feature vectors with the section morphological feature S to obtain a plurality of new state feature vectors, and merging the new state feature vectors into the abnormal sample set.

Respectively obtaining historical response values of continuous M historical excavation works according to the strong classifier, wherein M is a positive integer, and obtaining a response value threshold according to the historical response values: the state characteristic vector corresponding to the strong classifier with the response value larger than 0 belongs to the normal state characteristic vector, otherwise, the state characteristic vector corresponding to the strong classifier with the response value smaller than 0 belongs to the abnormal state characteristic vector; and calculating the mean value of the response values of Top-20% in the M historical response values, and taking the mean value of the response values as a threshold value of the response values.

The method comprises the following steps of respectively obtaining real-time response values of state characteristic vectors of real-time excavation work and response values of state characteristic vectors of K excavation works before the real-time excavation work by utilizing a strong classifier, wherein K is a positive integer, when the real-time response values and the response values are lower than response value thresholds, carrying out excavation abnormity early warning, and taking corresponding measures to carry out abnormity confirmation, and specifically comprises the following steps: when the response values of the K excavation works are not all smaller than the response value threshold value, the constructor checks the construction effect, but does not mean that a problem exists; and when the response values of the K excavation works are smaller than the response value threshold value, confirming that a problem occurs, and supervising and constructing staff need to manually evaluate the site.

The reason why the state monitoring of the excavation work is performed by using the Adaboost two classifier is that: the AdaBoost two-classification effect is a rough classification, due to the fact that the excavation conditions are variable and the operation characteristics of excavation, the change of the excavation section acquired after each excavation is carried out is continuous, the excavation working state is changed from a normal state to a critical state and then to an abnormal state, and therefore combined judgment is carried out based on continuous operation results, and the judgment result is more accurate.

In summary, the embodiment of the invention provides a municipal engineering excavation and filling auxiliary method based on excavation form measurement, the method obtains a compaction state stability index, a typical degree of an excavation volume increment value and a Fourier descriptor of excavation section form in excavation work so as to calculate the work special case degree of the excavation work, and the Fourier descriptor obtains the section form characteristics of the excavation work; forming a state feature vector of excavation work by the compaction state stability index, the typical degree of the excavation volume increment value, the special case degree of work and the section morphological feature; obtaining state feature vectors of a plurality of continuous excavation works to form a normal sample set and an abnormal sample set, training an Adaboost two-classifier by using the normal sample set and the abnormal sample set, and detecting abnormal states of the excavation works by using the trained Adaboost two-classifier. The invention can quickly and accurately carry out abnormity early warning so as to conveniently find problems in time and play an auxiliary role in excavation.

It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And that specific embodiments have been described above. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that are within the spirit of the present invention are intended to be included therein.

Claims (3)

1. A municipal engineering excavation and filling auxiliary method based on excavation form measurement is characterized by comprising the following steps:

taking a working process corresponding to a set excavation and filling path as excavation work, acquiring an excavation and filling compaction square coefficient and an excavation volume increment value in the current excavation work based on a set sampling frequency, and respectively and correspondingly obtaining an excavation and filling compaction square coefficient sequence and an excavation volume increment value sequence; calculating a compaction state stability index from the excavation and filling compactor coefficient sequence, and calculating a typical degree of excavation volume increment value from the excavation volume increment value sequence; acquiring the excavation section form of the current excavation work, and acquiring a Fourier descriptor of the excavation section form according to point cloud data of the excavation section form;

calculating the working special case degree of the current excavation work by combining the Fourier descriptor, the excavation and filling compaction square coefficient sequence and the compaction state stability index; taking the low-frequency component in the Fourier descriptor as the section morphological feature of the current excavation work; forming the compaction state stability index, the typical degree of the excavation volume increment value, the working special case degree and the section morphological characteristics into a state characteristic vector of the current excavation work;

acquiring the state characteristic vectors of a plurality of continuous excavation works, and clustering all the state characteristic vectors according to the difference value between the work special case degrees to obtain a normal sample set and an abnormal sample set; training an Adaboost two-stage classifier by using a normal sample set and an abnormal sample set to obtain a strong classifier of the Adaboost two-stage classifier, and detecting the abnormal state of excavation work by using the strong classifier;

the calculation formula of the compaction state stability index is as follows:

wherein, the first and the second end of the pipe are connected with each other,

is a stable index of a compaction state;

is a mean valueA function is transformed;

as a function of absolute value;

for the first of said sequence of fill and compaction square coefficients

Individual cut fill cube factor;

for the first in the sequence of fill and compaction square coefficients

Individual cut fill cube factor;

is a constant;

the method for acquiring the typical degree of the excavation volume increment value comprises the following steps:

respectively calculating the square of the difference between each excavation volume increment value in the excavation volume increment value sequence and the reference excavation volume to obtain the mean value of the square of the difference, and taking the product of the mean value and a correction coefficient as a difference index; the reference excavation volume refers to the excavation volume before compaction;

optimizing the difference index by using a hyperbolic tangent function to obtain the typical degree of the excavation volume increment value;

the calculation formula of the degree of the special working case is as follows:

wherein the content of the first and second substances,

is composed ofThe degree of the special case of the work; cosine () is a Cosine similarity function; mean () is the mean function; max () is a function for solving the maximum value;

is a reference fourier descriptor;

is the Fourier descriptor;

is the compaction state stabilization indicator;

filling the sequence of the compaction square coefficients for the excavation;

the method for acquiring the normal sample set and the abnormal sample set comprises the following steps:

respectively calculating the difference value between the special case degrees corresponding to the work in any two excavation works, taking the difference value as a sample distance, and carrying out DBSCAN clustering on all state feature vectors based on the sample distance to obtain a plurality of clustering clusters and isolated points;

and forming the state characteristic vectors corresponding to the isolated points into the abnormal sample set, and forming the state characteristic vectors in all the clustering clusters into the normal sample set.

2. The municipal works excavation and filling assistance method based on excavation form measurement as claimed in claim 1, wherein the abnormal sample set is optimized, and the optimization method comprises the following steps:

counting the working special case degrees in the abnormal sample set to obtain a median of the working special case degrees, obtaining target state feature vectors which are larger than the median in the abnormal sample set, randomly combining elements in all the target state feature vectors to obtain a plurality of new state feature vectors, and placing the new state feature vectors into the abnormal sample set.

3. The municipal works excavation and filling assisting method based on excavation form measurement as claimed in claim 1, wherein the method for detecting abnormal states of excavation work by using the strong classifiers comprises:

respectively obtaining historical response values of a plurality of historical excavation works according to the strong classifiers, and obtaining response value thresholds according to the historical response values;

and respectively acquiring a real-time response value of real-time excavation work and a response value of continuous K times of excavation work before the real-time excavation work according to the strong classifier, and performing excavation abnormity early warning when the real-time response value and the response value are lower than the response value threshold value.

Images